More Narcan Saves Lives, But Not Equally in Every County

A team of researchers at Penn State and elsewhere built a computer simulation of opioid addiction in six Pennsylvania counties — two big cities (Allegheny, which includes Pittsburgh, and Philadelphia), two mid-sized places (Erie, Dauphin), and two rural ones (Clearfield, Columbia). They then fast-forwarded to 2029 under three scenarios: a 10%, 20%, or 30% increase in naloxone distribution, and the same increases in buprenorphine prescribing. Naloxone is the nasal spray often sold as Narcan that can pull someone back from an overdose. Buprenorphine is a daily medication that reduces cravings and withdrawal, helping people stay in recovery.

The results were strikingly uneven. A 30% bump in naloxone was projected to reduce overdose deaths in Allegheny County by roughly 70% by 2029. The same percentage increase in rural Clearfield County produced about a 28% reduction, and in Erie County just 11%. Increasing buprenorphine helped too, but generally less — between 10% and 23% — with one exception: in Erie, buprenorphine beat naloxone, 20% versus 11%. The researchers traced the difference to each county's history. Places that already distribute naloxone widely and steadily tend to get more out of each additional dose, while places with erratic distribution see smaller returns.

That pattern is invisible if you only look at death tolls. Two counties can have nearly identical overdose death rates and need completely different responses. For local officials deciding how to spend limited public health money, this is the practical takeaway: the same policy is not equally powerful everywhere, and a strategy that works in Philadelphia may flop in a rural county. For families, it means the odds of a neighbor surviving an overdose depend heavily on where they happen to live and how well their county's addiction services already function.

The honest catch: this is a simulation, not a real experiment. No county actually changed its policies; the researchers projected what might happen using historical data and assumptions built into their model. Some of the projections come with very wide uncertainty — Clearfield's 28% estimate ranged anywhere from 2% to 64%. Real communities also face supply limits, stigma, pharmacy rules, and political resistance that a model cannot fully capture. The study points to where resources might do the most good, but it cannot promise those results will arrive.